AI Skills Hub - Browse 180,000+ Agent Skills

数据分析技能，提供结构化的数据查询、统计分析和报告生成工作流。当用户需要查询数据库、进行数据统计分析、生成分析报告时使用。按规范输出原始数据、计算结果、Markdown报告和HTML可视化报告。

Use when deploying Next.js projects to EdgeOne Pages - detects SSG/SSR project type, supports local CLI deployment or CNB pipeline, handles environment variables and region configuration (global/overseas).

Filter Amazon keywords by product image similarity using AI vision analysis with MCP (zai-mcp-server). Searches Amazon, merges product images into grids, then uses MCP to analyze visual features (color, style, material, shape) for intelligent matching. Use for filtering keywords from Seller Sprite Excel files, analyzing product similarity with AI instead of perceptual hashing, batch processing keywords with semantic visual understanding, and reducing MCP calls by merging images (10+ calls becomes 2 calls per keyword).

Agent Skill: CKEditor 5 development for TYPO3 v12+. Use when developing custom plugins, configuring RTE presets, or migrating from CKEditor 4. By Netresearch.

Creates and modifies Home Assistant Lovelace dashboards using the websocket API via Chrome DevTools. Use when the user wants to create, update, or delete Home Assistant dashboards, add cards or views, or query entities for dashboard building.

Convert legacy agent resources to the unified [Agent Skills](https://agentskills.io) format.

Use when importing web content (pages, YouTube videos, PDFs, audio) into local RAG for natural language queries. Trigger when user wants to save content for later querying or needs searchable knowledge base.

컨테이너 환경 검증 및 Dockerfile 빌드 테스트 스킬. 다음 상황에서 사용: (1) 사용자가 기능 개발 후 프로젝트가 컨테이너 환경에서 작동하는지 확인 요청할 때 (2) Dockerfile 빌드, 컨테이너 실행 테스트를 명시적으로 요청할 때 (3) 특정 컨테이너 런타임(docker, podman, nerdctl 등)으로 테스트 요청할 때 (4) 로컬 또는 원격 이미지로 컨테이너 실험을 원할 때 (5) Kubernetes/k3s/minikube 환경에서 배포 테스트할 때 지원 런타임: docker, podman, nerdctl, crictl, containerd 지원 오케스트레이션: kubectl, k3s, minikube, kind, lima

Convert weekly support metrics into human-readable narrative summaries for IT support managers. Use when users need to transform raw support ticket data (CSV/JSON format) into management-friendly markdown reports covering ticket volume trends, SLA performance, top issue categories, and key risks. Triggered by requests like "generate weekly support summary", "create support metrics report", "analyze support ticket data", or when working with support metrics files that need narrative explanation for leadership.

Convert C/C++ code into Strict SysY testcases using Python automation scripts.

Extract meeting recordings and transcripts from the Quill macOS app database. Outputs formatted markdown with speaker identification, transcripts, and metadata. Use when working with Quill meeting data.

Dynamic agent composition and management system. USE WHEN user says create custom agents, spin up custom agents, specialized agents, OR asks for agent personalities, available traits, agent voices. Handles custom agent creation, personality assignment, voice mapping, and parallel agent orchestration.

Scaffold a new hook registration. Keywords: hook, event, scaffolding.

Convert any file to markdown format using the markitdown library. Use this skill when users need to convert documents (PDF, DOCX, XLSX, PPTX, images, HTML, CSV, JSON, XML, audio files, etc.) into markdown format for easier reading, editing, or integration into markdown-based workflows.

Orchestrate long-running software development across multiple context windows using a two-agent architecture (initializer + coding agent). Use when starting new development projects, resuming work on existing projects, or implementing features incrementally. Enables consistent progress through structured artifacts (feature_list.json, claude-progress.txt, init.sh) and strict incremental workflows.

Automates `googleai_dart` updates when the Google AI OpenAPI specification changes.

Assists in creating Agent Skills of varying complexity levels (simple, moderate, complex). Use when the user wants to create, design, or build a new Agent Skill, or when they need guidance on skill architecture, workflow design, schema validation, or template structure.

Gravito 框架專用的靜態網站生成器。NOT a general-purpose SSG tool.

Difyアプリの作成・編集・検証・デプロイを自動化します。新規アプリ作成、既存アプリ修正、設定テストが必要な場合に使用してください。YAML生成・編集、検証、テスト実行をすべて自動で行います。

小红书内容智能发布工具。自动将文本整理成符合小红书格式的标题（≤20字）和正文，自动识别指定目录中的图片和视频文件，并通过xiaohongshu-mcp发布内容。适用于需要发布小红书内容、批量处理媒体文件、或自动化小红书运营的场景。

本 Skill 采用“Python 脚本 + references 文档”的官方推荐结构：`SKILL.md` 负责定义能力与调用入口，`scripts/` 提供可执行工具脚本，`references/` 收录 API/参数/合规等参考资料。

Phoenix LiveView patterns and real-time UI development. Use when implementing LiveView pages, handling PubSub events, building components, managing real-time updates, or working with dynamic interfaces.

Unity Editor operations via unity-mcp-client CLI. Handles build verification, console logs, tests, scene/object manipulation. Use after script edits or for debugging.

Orchestrate a 6-phase release workflow with automated service documentation, high-level docs updates, README "What's New" section, website improvements, and semantic versioning. Use when preparing a new release.

Manages Git worktrees for isolated parallel development. Handles creating, listing, switching, and cleaning up worktrees with automatic .env copying.

Use when you need to audit, fix, or improve an existing agent skill to meet Gold Standard compliance.

Scrape webpage content, extract metadata, download images, and prepare for import/migration to AEM Edge Delivery Services. Returns analysis JSON with paths, metadata, cleaned HTML, and local images.

Scrape webpage content, extract metadata, download images, and prepare for import/migration to AEM Edge Delivery Services. Returns analysis JSON with paths, metadata, cleaned HTML, and local images.

Collect market data from external sources with ethical scraping practices, proper rate limiting, and full provenance tracking.

Notion 데이터베이스에 GitHub README를 자동 동기화. 프로그래머스 코딩테스트 풀이를 GitHub에서 가져와 Notion 페이지로 생성. 한글 경로와 특수 공백 문자 처리 지원.

Python scripting with Gemini image generation using uv. Use when users ask to generate images, create AI art, edit images with AI, or run Python scripts with uv. Trigger phrases include "generate an image", "create a picture", "draw", "make an image of", "nano banana", or any image generation request.

Transcribe audio and video from URLs (YouTube, direct media links) using WhisperKit locally. Optionally analyze transcripts with AI when explicitly requested. Use when users provide URLs to media content and request transcription or speech-to-text conversion.

WSL2 Ubuntu administration from the Linux side. Profile-aware - reads preferences from the shared Windows-side profile at /mnt/c/Users/{WIN_USER}/.admin/profiles/{hostname}.json. Use when: apt packages, Docker containers, Python/uv in WSL, Node in WSL, shell configs (.bashrc/.zshrc), systemd services in WSL, SSH from WSL to servers. NOT for: .wslconfig changes (use admin-windows), Windows packages (use admin-windows), native Linux/macOS (use admin-unix). Keywords: wsl, wsl2, ubuntu, apt, docker, /mnt/c, wslpath, systemd wsl, python wsl, node wsl

This skill should be used when the user asks to "test Android app", "automate Android emulator", "tap button on Android", "take Android screenshot", "interact with Android UI", "ADB automation", "fill Android form", "swipe on emulator", "validate Android UI flow", or needs to control an Android device/emulator via ADB. Provides a unified CLI with JSON output optimized for LLM consumption.

Generate images using Google's Antigravity API (Gemini image generation). Use when user requests to create, generate, or make an image/picture (e.g., "generate a cat picture", "create an image of a sunset"). Credentials stored securely in ~/.config/antigravity/ (never in project directories).

基于 Saury.Revit.Template 创建 Revit 插件项目。当用户要求创建 Revit 插件、新建 Revit 项目、初始化 Revit 开发环境、使用 saury-revit 模板时触发。覆盖场景：(1) 创建新 Revit 插件项目 (2) 安装/更新 saury-revit dotnet 模板 (3) 在已创建的项目中添加新功能（Command/View/ViewModel/Service）。关键词：Revit、插件、plugin、saury-revit、dotnet new。

Integrate and operationalize the `cm-reporting` library in host React applications, including dependency setup, template delivery, component wiring, Snapshot import/export, Excel export, integrations callbacks, and legacy adapter flows. Use when users need to adopt this library, design production integration architecture, implement external pickers, troubleshoot runtime/export issues, or standardize delivery checklists.

Use this skill when users need to extract text from images, PDFs, or documents. Supports URLs and local files, with adaptive quality modes. Returns structured JSON containing recognized text, confidence scores, and quality metrics.

Retrieve and analyze test failures from Azure DevOps builds and Helix test runs for dotnet repositories. Use when investigating CI failures, debugging failing PRs, or given URLs containing dev.azure.com or helix.dot.net.

Swing trading system for US stocks using Colin Seow's TraderGPS strategy with MA crossover, volume confirmation, and systematic exits. Use when user says "scan stocks", "check positions", "backtest strategy", "stock alerts", or wants to run daily stock trading analysis. Monitors watchlist for buy/sell signals, tracks open positions, sends Telegram alerts, and provides backtesting capabilities.

新しいエージェントスキルの作成・編集・パッケージングを支援する。 ユーザーが「スキルを作りたい」「新しい能力を追加したい」「スキルを更新したい」 「知見をスキル化したい」と依頼した場合に使用する。 また、ユーザーとのやり取りで得た知見を蓄積する際にも使用する。 Agent Skills標準（agentskills.io）に準拠したスキルを作成する。

Creates ToolHive release PRs by analyzing commits since the last release, categorizing changes, recommending semantic version bump type (major/minor/patch), and triggering the release workflow. Use when cutting a release, preparing a new version, checking what changed since last release, or when the user mentions "release", "version bump", or "cut a release".

Edit and validate OpenClaw Gateway config (openclaw.json / JSON5). Use when adding/changing config keys (gateway.*, agents.*, models.*, channels.*, tools.*, skills.*, plugins.*, $include) or diagnosing openclaw doctor/config validation errors, to avoid schema mismatches that prevent the Gateway from starting or weaken security policies.

Render Mermaid diagrams as SVG and PNG using the Beautiful Mermaid library. Use when the user asks to render a Mermaid diagram.

Diagnose and fix claude-mem installation issues. Checks worker status, database integrity, service health, dependencies, and provides automated fixes for common problems.

Enforce focused, minimal, integrated implementations. Use when writing new features, reviewing code size, or verifying feature integration. Prevents over-engineering, orphan utilities, and verbose implementations.

使用 Playwright 与本地 Web 应用程序交互和测试的工具包。支持验证前端功能、调试 UI 行为、捕获浏览器截图和查看浏览器日志。必须通过 CDP 连接到外部已打开的 Chromium 浏览器进行测试。

Install and manage AI agent skills from Git repositories using the `npx skills add` command. Use this skill when users want to install skills from GitHub repositories or other Git sources, either installing a specific skill or all skills from a repository.

Diagnose and fix claude-mem installation issues. Checks PM2 worker status, database integrity, service health, dependencies, and provides automated fixes for common problems.

Multi-server SSH log collection agent for incident management

Extract authoritative property data from Maricopa County Assessor API including lot size, year built, garage spaces, pool, and valuations. Requires MARICOPA_ASSESSOR_TOKEN. Use for Phase 0 data collection, verifying listing claims, or obtaining kill-switch field values (lot, year, garage).

Calculate property scores using the 605-point weighted system (Location 250pts, Systems 175pts, Interior 180pts). Tier thresholds are UNICORN >=484, CONTENDER 363-483, PASS <363. Use when scoring properties, understanding tier assignments, analyzing score breakdowns, or running the scoring pipeline.

Download videos from YouTube and 1000+ other platforms with quality selection, format conversion, and metadata preservation. Use when users request to download, fetch, or save videos from URLs. Supports single videos, playlists, batch downloads, quality selection (480p-4K), multiple formats (MP4/WebM), audio extraction (MP3), and automatic metadata/thumbnail saving. Default output is ~/Downloads directory.

This skill guides through fixing validation failures following the System Delivery Playbook remediation procedures. Use when any validation script fails and you need structured remediation guidance.

Process and generate multimedia content using Google Gemini API for better vision capabilities. Capabilities include analyze audio files (transcription with timestamps, summarization, speech understanding, music/sound analysis up to 9.5 hours), understand images (better image analysis than Claude models, captioning, reasoning, object detection, design extraction, OCR, visual Q&A, segmentation, handle multiple images), process videos (scene detection, Q&A, temporal analysis, YouTube URLs, up to 6 hours), extract from documents (PDF tables, forms, charts, diagrams, multi-page), generate images (text-to-image with Imagen 4, editing, composition, refinement), generate videos (text-to-video with Veo 3, 8-second clips with native audio). Use when working with audio/video files, analyzing images or screenshots (instead of default vision capabilities of Claude, only fallback to Claude's vision capabilities if needed), processing PDF documents, extracting structured data from media, creating images/videos from text prompts, or implementing multimodal AI features. Supports Gemini 3/2.5, Imagen 4, and Veo 3 models with context windows up to 2M tokens.

项目知识库管理器 - 将保险业务项目文档(数据分析、PPT报告、技术方案)转化为结构化、可迁移的知识资产。支持:(1)自动生成2类标准化模板(项目启动/技术方案);(2)从Markdown/代码仓库智能提取技术决策和可复用模式;(3)统一文档格式标准;(4)生成可跨项目复用的知识库。触发场景:用户要整理项目文档、创建知识库、生成项目模板、提取技术经验、标准化现有文档时使用。

Diagnose and fix claude-mem installation issues. Checks PM2 worker status, database integrity, service health, dependencies, and provides automated fixes for common problems.

Deploy and manage Vercel projects. Use when deploying applications to Vercel, managing environment variables, checking deployment status, viewing logs, or performing Vercel operations. Supports production and preview deployments. Practical infrastructure operations - no "AI will build your app" magic.

React Native アプリケーションを複数のプラットフォーム（DeployGate、AppStoreConnect、GooglePlayConsole）にデプロイするスキル。テスト環境と本番環境を分離し、各環境に対応した環境変数とスキーマを参照します。

Process and generate multimedia content using Google Gemini API for better vision capabilities. Capabilities include analyze audio files (transcription with timestamps, summarization, speech understanding, music/sound analysis up to 9.5 hours), understand images (better image analysis than Claude models, captioning, reasoning, object detection, design extraction, OCR, visual Q&A, segmentation, handle multiple images), process videos (scene detection, Q&A, temporal analysis, YouTube URLs, up to 6 hours), extract from documents (PDF tables, forms, charts, diagrams, multi-page), generate images (text-to-image with Imagen 4, editing, composition, refinement), generate videos (text-to-video with Veo 3, 8-second clips with native audio). Use when working with audio/video files, analyzing images or screenshots (instead of default vision capabilities of Claude, only fallback to Claude's vision capabilities if needed), processing PDF documents, extracting structured data from media, creating images/videos from text prompts, or implementing multimodal AI features. Supports Gemini 3/2.5, Imagen 4, and Veo 3 models with context windows up to 2M tokens.

AI image generation and editing using nano-banana-2 API. Use this skill when users request to generate images from text prompts, edit existing images, or create visual content. Supports various aspect ratios (1:1, 4:3, 16:9, etc.) and quality levels (1K, 2K, 4K). Use for requests like "generate an image of...", "create a picture of...", "edit this image to...", or any visual content creation tasks.

Transform Unity Android Studio exports into Google Play Store ready AAB builds through systematic analysis and iterative problem-solving.

This skill should be used when the user asks to "extract Mermaid diagrams", "convert Mermaid to PNG", "generate Mermaid images", "create flowchart", "draw diagram", "设置主题", "change theme", "生成流程图", "转换图表", or mentions working with Mermaid code blocks in markdown files. Bundled scripts extract diagrams from markdown and generate high-quality PNG/SVG images with smart sizing and theme support.

Process and generate multimedia content using Google Gemini API for better vision capabilities. Capabilities include analyze audio files (transcription with timestamps, summarization, speech understanding, music/sound analysis up to 9.5 hours), understand images (better image analysis than Claude models, captioning, reasoning, object detection, design extraction, OCR, visual Q&A, segmentation, handle multiple images), process videos (scene detection, Q&A, temporal analysis, YouTube URLs, up to 6 hours), extract from documents (PDF tables, forms, charts, diagrams, multi-page), generate images (text-to-image with Imagen 4, editing, composition, refinement), generate videos (text-to-video with Veo 3, 8-second clips with native audio). Use when working with audio/video files, analyzing images or screenshots (instead of default vision capabilities of Claude, only fallback to Claude's vision capabilities if needed), processing PDF documents, extracting structured data from media, creating images/videos from text prompts, or implementing multimodal AI features. Supports Gemini 3/2.5, Imagen 4, and Veo 3 models with context windows up to 2M tokens.

Diagnose and fix claude-mem installation issues. Checks PM2 worker status, database integrity, service health, dependencies, and provides automated fixes for common problems.

Convert CSV files to JSON with support for custom delimiters, nested headers, quoted fields, and streaming

Generate realistic fake data from JSON schemas, producing 1-50 valid records that conform exactly to the schema. Use when creating sample datasets from schema definitions, populating test databases, or generating example API responses. Returns only valid JSON that passes schema validation.

Update VBW to the latest version with automatic cache refresh.

No description available

Use when creating, editing, visualizing, previewing, rendering, or debugging Mermaid diagrams, Graphviz DOT graphs, LaTeX for CS papers, or self-contained HTML files with rich visual assets (math, charts, 3D, code). Also use for scientific/engineering documents with Bode plots, WaveDrom timing diagrams, SVG circuit schematics, SI units, or professional technical document layout.

Build corpus-ready English TXT data from mixed file formats. Use when Codex needs to ingest raw text from PDF, TXT/Markdown, HTML/XML, DOCX, JSON/JSONL, CSV/TSV, or unknown text-like files; detect input formats at the start; clean and normalize extracted text; and produce presentable, analysis-ready corpus outputs.

Fetch and read Dart/Flutter package documentation from pub.dev and GitHub. Use when adding a new package dependency, learning how to use a package, checking package version/API, or needing package README documentation. Triggers on: "add package", "use package", "how to use X package", "package docs", "pub.dev", "package documentation", "install package", "flutter pub add", "dart pub add", any mention of integrating or using a specific pub.dev package.

Send emails with Markdown content, attachments, and multi-recipient support via SMTP. Use when Claude needs to send emails programmatically, including scenarios like: (1) Sending notifications or reports via email, (2) Sending Markdown-formatted content as HTML emails, (3) Sending emails with attachments, (4) Sending to multiple recipients with CC/BCC, (5) Configuring SMTP for different email providers (Gmail, QQ, 163, Outlook, etc.)

Use when the task requires automating a real browser from the terminal (navigation, form filling, snapshots, screenshots, data extraction, UI-flow debugging) via `playwright-cli` or the bundled wrapper script.

Browser automation with persistent page state and MetaMask wallet support. Use when users ask to navigate websites, fill forms, take screenshots, extract web data, test web apps, automate browser workflows, or interact with Web3 dApps. Trigger phrases include "go to [url]", "click on", "fill out the form", "take a screenshot", "scrape", "automate", "test the website", "log into", "connect wallet", "sign transaction", "MetaMask", "dApp testing", or any browser/Web3 interaction request.

OpenClaw EC2 인스턴스에 expert-team 페르소나를 배포한다. "/deploy", "배포해줘", "OpenClaw 띄워줘", "페르소나 배포", "{역할} 띄워줘", "인스턴스 생성해줘", "deploy status" 요청 시 사용한다. 페르소나별 SOUL.md/IDENTITY.md/AGENTS.md를 관리하고, Pulumi IaC로 AWS EC2에 배포한다.

リスト（Excel/CSV）とWebページを紐付けて詳細情報を自動取得するスクレイピングツール。 使用タイミング: (1) 商品リストから商品詳細ページの情報を取得したい (2) URLリストから各ページの特定要素を抽出したい (3) IDリストとURL規則からWebページをバッチ取得したい (4) 大量のWebページから同一フォーマットの情報を収集したい キーワード: スクレイピング、Webスクレイピング、データ抽出、リスト取得、バッチ取得、Excel、CSV

智能摄影学习助手，支持照片构图美学诊断、评分拆解、个性化学习建议、定制化练习方案和情感解读

Tag-based upstream sync for a fork/secondary-development repo: fetch tags, let the user choose a tag, create a fresh branch from that tag, then read the old customization branch’s git changes + intent Markdown to re-implement the requirements on the new branch (no merge/rebase of the old branch).

End-to-end AI video production with avatar narration (HeyGen) and programmatic composition (Remotion). Use for creating tutorials, explainers, product demos, and narrated videos. Supports multi-language, custom branding, screen recordings, and batch clip generation.

Manages Printify print-on-demand operations via the Printify Public API (V1 + V2 Catalog). Supports shops, catalog exploration, products, orders, events, webhooks, and rate-limit-aware workflows using OAuth 2.0 or Personal Access Tokens.

Deploy and manage Vercel projects. Use when deploying applications to Vercel, managing environment variables, checking deployment status, viewing logs, or performing Vercel operations. Supports production and preview deployments. Practical infrastructure operations - no "AI will build your app" magic.

Use the Fly.io flyctl CLI for deploying and operating apps on Fly.io: deploys (local or remote builder), viewing status/logs, SSH/console, secrets/config, scaling, machines, volumes, and Fly Postgres (create/attach/manage databases). Use when asked to deploy to Fly.io, debug fly deploy/build/runtime failures, set up GitHub Actions deploys/previews, or safely manage Fly apps and Postgres.

Query and monitor Unraid servers via the GraphQL API. Use when the user asks to 'check Unraid', 'monitor Unraid', 'Unraid API', 'get Unraid status', 'check disk temperatures', 'read Unraid logs', 'list Unraid shares', 'Unraid array status', 'Unraid containers', 'Unraid VMs', or mentions Unraid system monitoring, disk health, parity checks, or server status.

Personal knowledge base with SQLite + FTS5. Index contacts, documents, ChatGPT exports, and WhatsApp data. Query everything instantly with full-text search. Use for contact lookups, conversation history search, document retrieval, and building persistent memory systems.

Crypto Trading | Crypto Search | Crypto Token Filter -- AI crypto trading wallet via natural language. Use when the user wants to execute AI research tasks, trade crypto autonomously, search crypto content/news, filter projects by market criteria, manage trading positions, or interact with DROYD agents. Supports agent tasks (research, trading, data analysis), content search (semantic/recent), project discovery (by name/symbol/address/concept), project filtering (market cap, momentum, technical indicators), watchlist management, and autonomous trading with stop losses, take profits, and quant-based strategies. Works with Solana (trading) and other chains Ethereum, Base, and Arbitrum for token filtering + research.

Monitor SEC EDGAR for new filings and get Telegram/Slack summaries via Clawdbot. Use when setting up SEC filing alerts, adding/removing tickers to monitor, configuring form types, starting/stopping the watcher, or troubleshooting filing notifications.

Manage documents in Paperless-ngx - search, upload, tag, and retrieve.

Search the web using SearXNG. Use when you need current information, research topics, find documentation, verify facts, or look up anything beyond your knowledge. Returns ranked results with titles, URLs, and content snippets.

Fetch Chinese stock and futures market data via Tushare API. Supports stock quotes, futures data, company fundamentals, and macroeconomic indicators. Use when the user needs financial data from Chinese markets. Requires TUSHARE_TOKEN environment variable.

Analyze CI build and test status from Azure DevOps and Helix for dotnet repository PRs. Use when checking CI status, investigating failures, determining if a PR is ready to merge, or given URLs containing dev.azure.com or helix.dot.net.

Analyze VMR codeflow PR status for dotnet repositories. Use when investigating stale codeflow PRs, checking if fixes have flowed through the VMR pipeline, or debugging dependency update issues in PRs authored by dotnet-maestro[bot].

Document conversion routing with dual invocation support

Interactive sub-agent creation skill for Claude Code. Use when user wants to create a custom subagent or mentions needing a specialized agent for specific tasks. This skill guides the entire subagent creation process including identifier design, system prompt generation, skill/context selection, and writing properly formatted agent files to .claude/agents.

Advanced analytics and insights for Ralph orchestration diagnostics. Use when analyzing Ralph loop performance, detecting anomalies, comparing sessions, generating visual dashboards, investigating backpressure issues, debugging stuck loops, or performing post-mortem analysis on Ralph diagnostic data. Triggers include requests to "analyze Ralph diagnostics", "generate dashboard", "compare sessions", "find anomalies", "visualize Ralph data", or working with .ralph/diagnostics/ directories.

Dify Workflow DSL 자동생성 (STEP 2)

Build and evolve NestJS backends with correct DI/module boundaries, request lifecycle hygiene (pipes/guards/interceptors/filters), validation + serialization, OpenAPI, and testing. Use when adding or refactoring NestJS endpoints/modules and needing deterministic steps, output contracts, and verification gates.

Verifies UI changes via rigorous, evidence-based visual validation (screenshots/video/URLs) to catch regressions, design-system drift, responsive breakage, and visual accessibility issues.

Provides high-signal, fast code review with selectable modes (quality, security, performance, tooling). Includes an optional safe-by-default review script to summarize diffs, scan for risky patterns, and produce a deterministic report.

Work with PDFs safely and repeatably: extract text/tables, convert pages to images, inspect/fill forms, and produce verifiable outputs (markdown/json/images/filled pdf). Use when a task involves PDF documents.